ACDT #103 meets at 21:00 tonight to present audit findings on the Glamsterdam repricing EIPs 8037 and 8038. Vitalik reposted a call for app developers to join, and Ethlabs recommends no decision until the impact is clear. On post-quantum accounts, ethresear.ch got its first app-side measurements: a Safe that only accepts hash-based signatures works with Permit2 but not with EIP-2612 permit, and ERC-8499 now derives the mnemonic fingerprint from a public key. The rest of Sunday was mostly ERC status moves and debate over the pqai27 timeline.
Last updated at 18:02 on 2026-11-15. The call is at 21:00 on 11-16 (14:00 UTC). MariusVanDerWijden opens with the current status; ChainSecurity and Dedaub each present independent audit findings; nerolation covers the EF's internal investigation; and EF EVM Dev Tooling reports on app outreach, including Juicebox protocol compatibility concerns and PoCs in mejango/glamsterdam-pocs. The stated goal is a shared understanding of the findings, their impact on deployed contracts and what's still uncertain, to inform Thursday's ACDE.
ACDT #103 agenda Last updated at 18:02 on 2026-11-15. The call is at 21:00 on 11-16 (14:00 UTC). MariusVanDerWijden opens with the current status; ChainSecurity and Dedaub each present independent audit findings; nerolation covers the EF's internal investigation; and EF EVM Dev Tooling reports on app outreach, including Juicebox protocol compatibility concerns and PoCs in mejango/glamsterdam-pocs. The stated goal is a shared understanding of the findings, their impact on deployed contracts and what's still uncertain, to inform Thursday's ACDE.
At 12:19 on 2026-11-15: app developers are asked to join Monday's call with core devs and auditors to gather evidence on how EIP-8037/8038 affect deployed contracts.
At 07:20 on 2026-11-16: Ethlabs has spent more than two weeks monitoring, triaging and analysing reports (formal ones and informal ones posted on X) about EIP-8037/8038's impact on deployed contracts, and recommends no Glamsterdam decision until there is clarity. Details are in Tracking the Glamsterdam discussion.
Caspar's Ethlabs weekly update At 07:20 on 2026-11-16: Ethlabs has spent more than two weeks monitoring, triaging and analysing reports (formal ones and informal ones posted on X) about EIP-8037/8038's impact on deployed contracts, and recommends no Glamsterdam decision until there is clarity. Details are in Tracking the Glamsterdam discussion.
Which contract patterns tonight's audits flag, and whether any mitigation is proposed, such as versioned opcodes. EIP-8488, versioned opcodes for gas repricing, listed in Ethereal news #47 (05:08 on 2026-11-16), takes exactly that approach.
At 18:13 on 2026-11-15, johba reported on a Sepolia Safe whose owners only accept hash-based signatures, so no ECDSA key can sign for it, and ran it through the permit flows:
At 16:05 on 2026-11-15 the author johba replied that the fingerprint is now fixed as the first 4 bytes of SHA-256 over account 0's SLH-DSA-SHA2-128s public key (8 hex digits), and is never computed from the root or a seed directly. On a server CPU that key generation took 72 ms against 3.6 ms for PBKDF2, so each guess against a written-down fingerprint costs about 20 times more than PBKDF2 alone. Showing it stays a SHOULD. For stateful keys, the draft now requires a generator seeded from a source that doesn't repeat across restarts, and says a device that can't keep state reliably must not hold a stateful instance. The changes are in ERCs PR #2172 (updated 18:10 on 2026-11-15).
ERC-8499 discussion At 16:05 on 2026-11-15 the author johba replied that the fingerprint is now fixed as the first 4 bytes of SHA-256 over account 0's SLH-DSA-SHA2-128s public key (8 hex digits), and is never computed from the root or a seed directly. On a server CPU that key generation took 72 ms against 3.6 ms for PBKDF2, so each guess against a written-down fingerprint costs about 20 times more than PBKDF2 alone. Showing it stays a SHOULD. For stateful keys, the draft now requires a generator seeded from a source that doesn't repeat across restarts, and says a device that can't keep state reliably must not hold a stateful instance. The changes are in ERCs PR #2172 (updated 18:10 on 2026-11-15).
Opened by Marchhill at 19:09 on 2026-11-15. It defines the generic verification key identifier as a domain-separated hash of a 32-byte program commitment, split out of EIP-8500 for separate review. At the same time (19:10) the EIP-8500 PR got a commit delegating committed verification keys to EIP-8288 and now lists #12667 as a dependency.
EIP-8288 amendment PR: committed verification keys Opened by Marchhill at 19:09 on 2026-11-15. It defines the generic verification key identifier as a domain-separated hash of a 32-byte program commitment, split out of EIP-8500 for separate review. At the same time (19:10) the EIP-8500 PR got a commit delegating committed verification keys to EIP-8288 and now lists #12667 as a dependency.
At 19:01 on 2026-11-15, RealSpap (who runs a related project and disclosed possible bias) shared a measurement: the same Safe address 0xcBa28b38103307Ec8dA98377ffF9816C164f9AFa has 10 signers with a threshold of 6 on Ethereum mainnet, and 9 signers with a threshold of 5 on Base mainnet, with replayable cast commands. The takeaway: one shared address doesn't mean the same signers.
A new reply at 18:59 on 2026-11-15 discussed test cases. The draft (opened 2026-11-07) adds an optional "protected" flag to ERC-20s: small sends stay instant, larger ones are queued and can be cancelled by a pre-registered backup wallet, and a "stress" action sweeps the balance to the backup and locks it, which helps against stolen keys and coercion.
Merged at 03:56 on 2026-11-16, and ERC-8328 moved to Final at 03:49. Both are append-only "subject-linked" record standards (a compliance event log and an impact snapshot log). From the same series, ERC-8330, a NAV snapshot oracle, moved to Last Call at 00:29 on 2026-11-16, with a last-call-deadline of 2026-11-30 in the file. Teams putting RWA or fund NAV data on-chain can start integrating.
ERC-8329 moved to Final merged at 03:56 on 2026-11-16, and ERC-8328 moved to Final at 03:49. Both are append-only "subject-linked" record standards (a compliance event log and an impact snapshot log). From the same series, ERC-8330, a NAV snapshot oracle, moved to Last Call at 00:29 on 2026-11-16, with a last-call-deadline of 2026-11-30 in the file. Teams putting RWA or fund NAV data on-chain can start integrating.
ERC-8329 moved to Final merged at 03:56 on 2026-11-16, and ERC-8328 moved to Final at 03:49. Both are append-only "subject-linked" record standards (a compliance event log and an impact snapshot log). From the same series, ERC-8330, a NAV snapshot oracle, moved to Last Call at 00:29 on 2026-11-16, with a last-call-deadline of 2026-11-30 in the file. Teams putting RWA or fund NAV data on-chain can start integrating.
Updated at 08:10 on 2026-11-16: the author reproduced an NFT that reports 1 ETH of backing while a 0.01 ETH reservation blocks full redemption, and proposes that strict ERC-8060 valueOf must match what can actually be redeemed, with marketplaces checking reservation state atomically at settlement. When buying these NFTs, don't rely on the value shown at listing time.
At 09:28 on 2026-11-15: the spec now covers several deployed formats OpenSea already handles (raw JSON, decimal {id} templates and more), and clients should refetch media when a metadata update event fires.
Posted at 21:29 on 2026-11-15, followed at 22:40 by a third Sepolia evidence run (58 transactions, all successful). Agent ERCs are still changing fast, so pin a version before integrating.
Meets at 23:00 on 11-18 (agenda updated 02:00 on 2026-11-16) to discuss AI in the EIP process and exceptional merges after two weeks, among other items.
At 05:08 on 2026-11-16: EIP-8487 separate execution and billing credits, EIP-8488 versioned opcodes, EIP-8491 LUCID key publisher priority lists, EIP-8496 code reads in block level access lists, and EIP-8497 FOCIL omission witnesses for Frame transactions.
At 07:20 on 2026-11-16 expects the next few ACD calls to make important decisions on the direction of upcoming hard forks, including whether to reopen Hegotá decisions to speed up pqai timelines. The Ethlabs account shared it at 07:22.
Caspar on pqai27 and ETH as a store of value at 07:20 on 2026-11-16 expects the next few ACD calls to make important decisions on the direction of upcoming hard forks, including whether to reopen Hegotá decisions to speed up pqai timelines. The Ethlabs account shared it at 07:22.
At 00:50 on 2026-11-16: "The PQ timeline of 2028 already felt like a big acceleration, and pulling it even sooner seems borderline too fast". The original is from mikeneuder at 23:02 on 2026-11-15, responding to the post-quantum pieces from potuz and Ethlabs.
Lightclients quoting Mike Neuder at 00:50 on 2026-11-16: "The PQ timeline of 2028 already felt like a big acceleration, and pulling it even sooner seems borderline too fast". The original is from mikeneuder at 23:02 on 2026-11-15, responding to the post-quantum pieces from potuz and Ethlabs.
At 18:38 on 2026-11-15: CelticWarrior backs prioritizing post-quantum but worries about solo staking, whose share he sees declining at an accelerating pace.
At 13:01 on 2026-11-15. Affordable hardware scales more slowly than Ethereum needs, and while compute and PQ have solutions, state has only seen baby steps.